Agents in the Org Chart:​ AI Agent Accountability and Supervision Roundtable​

Singapore's public sector has moved fast on AI adoption — but a new phase has begun. With IMDA's Model AI Governance Framework for Agentic AI now in place and open questions on legal responsibility for AI agents still being worked through, agencies are no longer asking whether to deploy AI agents, but how to run them responsibly once they're part of daily operations. ​

​

For agencies handling national security, patient care, land use, transport safety, financial regulation, and citizen trust,this is not a hypothetical: AI agents are already touching real decisions, real data, and real accountability chains. This roundtable brings together leaders across defence, health, urban development, transport, finance, home affairs, and digital government to work through the practical, operational questions that policy frameworks alone don't answer — how to supervise AI agents day to day, how officers and AI should actually divide and share work, how to assign accountability when something goes wrong, and what structure is needed to manage this at the scale of a whole public service, not a single pilot.​

 

Key Takeaways:

• Define Your AI Red Lines - Identify which decisions can be delegated to AI agents and which must remain under human control.​
• Learn from Peer Agencies - Explore how public sector organisations are approaching supervision, accountability and workforce readiness for agentic AI.​
• Build an Operational Governance Model - Understand what processes, roles and platforms are needed to move beyond pilots and deploy AI agents at scale.​

​